Re: [idvusers] Formating NetCDF headers for Point Data

William and Don,

You can put many time steps and many variables and many locations in one point data file for the IDV, either in NetCDF or in ASCII format. Here are some examples of point data for the IDV. These were made for the GEON IDV and may not display fully in the IDV, but that is a matter of special symbols for geophysical data. The IDv should be able to read all of them and make some kind of display for all the variables and times.

This file has a nice example of global attributes using the "NetCDF Attribute 
Convention for Dataset Discovery (DRAFT)"
( This file should have complete dates and times, not year numbers, but it works.

2. use the IRIS Earthquake Browser ( The IRIS Earthquake catalog has more than 1,900,000 world-wide earthquake locations with depths and magnitudes. The Earthquake Browser uses a Google Maps web interface to the IRIS hypocenter catalog. Open the IRIS Earthquake Browser in a new window. It uses a GoogleMap interface. Select an area of interest on the map, and optionally, how many hypocenters you want, and time and magnitude ranges. The Earthquake Browser finds the data that matches your search. A URL link to a NetCDF file is provided , at the "NetCDF" link in the top title bar, between "Table" and "XML."

3. an earthquake cdl file:
This is a nice example, except time is in the clumsy "seconds since 1970" 

4. here is a cdl file with stations 4 character ids in the variable list (but 
missing from the data):

5. In many cases I find the ascii .csv file format convenient for point data 
with stations and times; see
"Point Data in Text Format (csv files)"  
A good example with station names (IDN) is
Notice that complete dates and times to the second are given.

6. from the web page "Track Data for the GEON IDV"
aircraft track:
animal track:
Track data is a point data set with a moving sequence of points.

For point data a precise time description is always a question. A way to put time in a format like "yyyy-MM-dd HH:mm:ss" in a NetCDF file would be useful. It may be possible and I simply don't know. Time to the millisecond or microsecond may soon be useful.

Stuart Wier


I'm trying to package some point data for IDV using netCDF.

Are there any examples of setting this up except for

Specifically what we want to do is take a days worth of point data at
different stations and collect them in a single file for use in IDV (and
other tools)
e.g., for a given parameter
float temperature(station,time) ; Would such a setup work or am I limited one time step per file as the
madis examples seem to imply.